def __init__(self,movieID):

self.movieID = movieID

self.path = './movieproject/{}'.format(movieID)

# 处理城市名称不一致问题

# 猫眼的城市名不带有‘市’‘县’字样，而地图数据是带的，例如烟台市在猫眼显示为‘烟台’

def city_handles(self,cities):

with open(r'./venv/lib/python3.6/site-packages/pyecharts/datasets/city_coordinates.json','r',

encoding='utf-8') as f:

data = json.loads(f.read())

data_new = data.copy()

for city in cities:

# 处理城市为空的数据

if city == '香格里拉':

print(city)

if city == '':

cities.remove(city)

break

for k in data:

if k == city:

break

if k.startswith(city):

# print('k=%s,city=%s'%(k,city))

data_new[city] = data[k]

# print('新增地图数据%s' % city)

break

if k.startswith(city[0:-1]) and len(city)>2:

data_new[city] = data[k]

# print('新增地图数据%s' % city)

break

#处理不存在于地图库的地名

if city not in data_new:

# print('%s不在地图库里' % city)

cities.remove(city)

print('地图库数据%s条'% len(data_new))

# 覆盖到地图库的数据文件

with open(r'./venv/lib/python3.6/site-packages/pyecharts/datasets/city_coordinates.json','w',

encoding='utf-8') as f:

f.write(json.dumps(data_new,ensure_ascii=False))

return cities

# 从文件读取数据,返回dict类型

def read_csv(self,path='default'):

if path == 'default':

path = self.path + r'/{}.csv'.format(self.movieID)

time = []

city = []

gender = []

nickname = []

userlevel = []

score = []

content = []

data_dict={

'time':time,

'city':city,

'gender':gender,

'nickname':nickname,

'userlevel':userlevel,

'score':score,

'content':content

}

with open(path,'r',encoding='utf-8',newline='') as f:

reader = csv.reader(f)

i = 0

for row in reader:

if i != 0:

time.append(row[0])

city.append(row[1])

gender.append(row[2])

nickname.append(row[3])

userlevel.append(row[4])

score.append(row[5])

content.append(row[6])

i +=1

print('共导入评论数据%s条' % i)

return data_dict

# 分析等级 和 评分，生成饼中饼图 和 玫瑰饼图

def score_distribution(self,levels,scores):

pie_l = Pie('评论者等级饼图')

attr_l = ['一级','二级','三级','四级','五级']

v_l = []

v_l.append(levels.count('1'))

v_l.append(levels.count('2'))

v_l.append(levels.count('3'))

v_l.append(levels.count('4'))

v_l.append(levels.count('5'))

pie_l.add(

"",

attr_l,

v_l,

radius=[60, 75],

is_random=True,

is_label_show=True

)

pie_l.add(

"",

attr_l,

v_l,

radius=[0, 55],

rosetype="area",

)

# pie_l.render(self.path +'/level_pie.html')

scores_coun = Counter(scores)

scores_list = sorted(scores_coun.items(), key=lambda x: x[0], reverse=False)

# print(scores_list)

attr_s = []

v_s =[]

for score in scores_list:

attr_s.append(score[0] + '分')

v_s.append(score[1])

pie_s = Pie('评论者评分饼图')

pie_s.add(

"",

attr_s,

v_s,

is_random=True,

radius=[30, 78],

rosetype="area",

is_label_show=True,

legend_orient='vertical',

legend_pos='left',

legend_top='center'

)

page = Page()

page.add(pie_l)

page.add(pie_s)

page.render(self.path +'/levelscore_pie.html')

# 分析性别，生成饼状图

def gender_distribution(self,gender):

list_num = []

list_num.append(gender.count('0'))

list_num.append(gender.count('1')) #男数量和

list_num.append(gender.count('2')) #女数量和

attr = ['未知','男','女']

pie = Pie('性别比例图')

pie.add('',attr,list_num,is_label_show =True)

pie.render(self.path + r'/gender_pie.html')

# 分析评论时间，生成时间折线图

def time_distribution(self, times):

# 计算v1，按小时统计,计算V2,按天统计

date_hour = []

date_day = []

for time in times:

dtime = datetime.datetime.strptime(time, '%Y-%m-%d %H:%M:%S')

date_hour.append(dtime.strftime('%m-%d|%H'))

date_day.append(dtime.strftime('%m-%d'))

# 按时间升序排序

date_hour.sort()

date_day.sort()

# 不能用counter类，因为他会排序

date_hour_dict = {}

for date in date_hour:

if date not in date_hour_dict:

date_hour_dict[date] = 1

else:

date_hour_dict[date] += 1

# print(date_hour_dict)

# 每日总数

date_day_dict = {}

for date in date_day:

if date not in date_day_dict:

date_day_dict[date] = 1

else:

date_day_dict[date] += 1

# print(date_day_dict)

line_top = Line("评论时间(小时)折线图")

attr1, v1 = line_top.cast(date_hour_dict)

# print(attr)

# print(v)

line_top.add(

"时间(小时)",

attr1,

v1,

mark_point=['max'],

mark_line=['average'],

is_fill=True,

area_color="#FFCCCC",

area_opacity=0.3,

is_smooth=True,

is_datazoom_show=True,

datazoom_type="both",

datazoom_xaxis_index=[0, 1]

)

# line.render(self.path + r'/time_line.html')

line_bottom = Line("")

attr2, v2 = line_bottom.cast(date_day_dict)

line_bottom.add(

"时间(天)",

attr2,

v2,

legend_pos="60%",

mark_point=["max", "min"],

mark_line=["average"],

# is_yaxis_inverse=True,

xaxis_pos="top",

is_datazoom_show=True

)

grid = Grid(width=1200, height=800)

grid.add(line_top, grid_bottom='60%')

grid.add(line_bottom, grid_top='50%')

grid.render(self.path + r'/time_line.html')

# 分析城市，生成可视化地图分布

def city_distribution_bar(self, cities):

# 统计城市名称,该城市出现的次数

# for city in cities:

# if city not in city_dict:

# city_dict[city] = cities.count(city)

# 按照次数大小排序

city_dict = Counter(cities)

sort_list = sorted(city_dict.items(),key=lambda x:x[1],reverse=True)

city_name = []

city_num = []

for i in range(len(sort_list)):

city_name.append(sort_list[i][0])

city_num.append(sort_list[i][1])

bar = Bar('评论者城市图示')

bar.add('',city_name,city_num,

is_datazoom_show=True,

datazoom_type='both',

is_label_show=True,

)

bar.render(self.path + r'/city_bar.html')

def city_distribution_geo(self, cities):

# 统计城市名称,该城市出现的次数

for city in cities:

if city == '':

cities.remove(city)

coun = Counter(cities)

# 处理地图库数据

city_name = self.city_handles(list(coun))

city_dict = dict(coun)

city_num = []

for city in city_name:

city_num.append(city_dict[city])

# print('%s : %s' %(len(city_name), city_name))

# print('%s : %s' %(len(city_num), city_num))

city_max = max(city_num)

geo = Geo('评论者城市分布情况',

"data from maoyan",

title_color="#fff",

title_pos="center",

width=1200,

height=600,

background_color="#404a59"

)

geo.add('',city_name,city_num,

type='scatter',

maptype='china',

visual_range=[0, city_max],

visual_text_color="#fff",

symbol_size=15,

is_visualmap=True

# is_piecewise=True

# visual_split_number=10

)

geo.render(self.path + r'/city_geo.html')

# 利用评论内容，生成词云图

def content_distribution(self,contents):

jieba.load_userdict('userdict.txt')

content_cut = jieba.cut(str(contents),cut_all=False)

words = ' '.join(content_cut)

# 设置屏蔽词

# stopwords = STOPWORDS.copy()

# stopwords.add('电影')

stopwords = set()

with open('stopwords.txt','r',encoding='utf-8') as f:

rows = f.readlines()

for row in rows:

stopwords.add(row.replace('\n',''))

# print(stopwords)

#导入背景图

bg_image = plt.imread('bgimage.jpg')

# 设置词云参数，参数分别表示：画布宽高、背景颜色、背景图形状、字体、屏蔽词、最大词的字体大小

wc = WordCloud(width=1200,height=800,background_color='white',mask=bg_image,

font_path='STKAITI.ttf',stopwords=stopwords,max_font_size=300,

random_state=40)

# 分词传入云图

wc.generate_from_text(words)

plt.imshow(wc)

# 不显示坐标轴

plt.axis('off')

plt.show()

wc.to_file(self.path+r'/wordcloud.jpg')

def all_distribution(self):

# 加载评论数据

data_dict = self.read_csv(path='default')

# 性别饼图

self.gender_distribution(data_dict['gender'])

# 城市柱状图

self.city_distribution_bar(data_dict['city'])

# 地理坐标系

self.city_distribution_geo(data_dict['city'])

# 评论时间折线图

self.time_distribution(data_dict['time'])

# 生成评论等级和评分饼图

self.score_distribution(data_dict['userlevel'], data_dict['score'])

# 评论内容生成词云图